key: cord-0758499-e9gaeckg authors: Nguyen, Jennifer L; Alfred, Tamuno; Reimbaeva, Maya; Malhotra, Deepa; Khan, Farid; Swerdlow, David; Angulo, Frederick J title: Population attributable fractions of underlying medical conditions for COVID-19 diagnosis and COVID-19 hospitalizations, ventilations, and deaths among adults in the United States date: 2022-03-24 journal: Open Forum Infect Dis DOI: 10.1093/ofid/ofac099 sha: 11e76b7aa338ce785dc6ccbd9fa9486b2860bb12 doc_id: 758499 cord_uid: e9gaeckg BACKGROUND: Several underlying medical conditions have been reported to be associated with an increased risk of COVID-19 disease, hospitalization, and death. Population attributable fractions (PAFs) describing the proportion of disease burden attributable to underlying medical conditions for COVID-19 diagnosis and outcomes have not been reported. METHODS: A retrospective population-based cohort study was conducted using Optum’s de-identified Clinformatics (®) Data Mart database. Individuals were followed from January 20, 2020 – December 31, 2020 for diagnosis and clinical progression, including hospitalization, intensive care unit admission, intubation and mechanical ventilation or extracorporeal membrane oxygenation, and death. Adjusted rate ratios and PAFs of underlying medical conditions for COVID-19 diagnosis and disease progression outcomes were estimated by age (years; 18-49, 50-64, 65-74, ≥75), sex, and race/ethnicity. RESULTS: Of 10,679,566 cohort members, 391,964 (3.7%) were diagnosed with COVID-19, of whom 87,526 (22.3%) were hospitalized. Of those hospitalized, 26,640 (30.4%) died. Overall, cardiovascular disease and diabetes had the highest PAFs for COVID-19 diagnosis and outcomes of increasing severity across age groups (up to 0.49 and 0.35, respectively). Among adults ≥75, neurologic disease had the second highest PAFs (0.05‒0.27) after cardiovascular disease (0.26‒0.44). PAFs were generally higher in Black persons than in other race/ethnicity groups for the same conditions, particularly in the two younger age groups. CONCLUSIONS: A substantial fraction of the COVID-19 disease burden in the US is attributable to cardiovascular disease and diabetes, highlighting the continued importance of COVID-19 prevention (e.g., vaccination, mask wearing, social distancing) and disease management of patients with certain underlying medical conditions. The coronavirus disease 2019 pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has resulted in more than 396 million reported cases and >5.7 million deaths worldwide, including >75 million cases and >895,000 deaths in the United States, as of February 8, 2022 [1] . Older persons, males, and persons with underlying medical conditions are at increased risk of COVID-19 diagnosis and COVID-19 disease progression including medical complications and death [2] [3] [4] [5] [6] . Underlying medical conditions associated with an increased risk of severe COVID-19 outcomes include cardiovascular disease, chronic obstructive pulmonary disease (COPD), hypertension, diabetes, chronic kidney disease, obesity, and cancer [3, 4, 7, 8] . Among patients with COVID-19, the presence of cardiovascular disease, diabetes, or hypertension increases the relative risk of progression to severe disease by approximately 1.5, and the relative risk of death by 1.5-2.1 [4] . Describing the COVID-19 disease burden attributable to underlying medical conditions can help focus COVID-19 prevention, treatment, and management of persons with underlying medical condition [9, 10] . The population attributable fraction (PAF) is an epidemiologic tool that estimates the proportion of disease burden that is attributable to a particular risk factor [11, 12] . PAFs of underlying medical conditions for COVID-19 diagnosis and disease outcomes have not been reported. The aim of this study was to quantify and compare the burden of COVID-19 attributable to selected underlying medical conditions in the United States. A population-based study was conducted using Optum's de-identified Clinformatics ® Data Mart (CDM) database. CDM is derived from a database of administrative health claims for members of A c c e p t e d M a n u s c r i p t 5 large commercial and Medicare Advantage health plans from all 50 US states. Administrative claims submitted for payment by providers and pharmacies are verified, adjudicated and de-identified before inclusion in CDM. These data, including patient-level enrollment information, are derived from claims submitted for healthcare services at outpatient medical clinics, pharmacies, hospital emergency departments, and inpatient hospital settings. Medical conditions are identified using diagnosis and procedure codes attached to individual claims. Death information is available as year and month of death only. A retrospective cohort was created that included all adults age ≥18 years old in the CDM database during the time period January 20, 2019 -January 19, 2020 provided they did not have any gaps in insurance coverage of >45 days during this one-year baseline period, as recommended in the US Food and Drug Administration (FDA) Sentinel Initiative COVID-19 Natural History Master Protocol [13] . This cohort was followed from January 20, 2020 (i.e., the date of the first laboratory-confirmed SARS-CoV-2 infection in the United States) to December 31, 2020 for COVID-19 diagnosis and COVID-19 disease progression outcomes. Demographic characteristics of interest included age (in 2019), sex, and race/ethnicity. Age was stratified into 4 categories: 18-49 years, 50-64 years, 65-74 years, and ≥75 years. As stated above, administrative claims submitted for payment by providers and pharmacies are Cohort members were classified as having ≥1 of the following underlying medical conditions if they had ≥2 reports of that condition on separate health encounters during January 20, 2019 -January 19, 2020: asthma; cardiovascular disease; cerebrovascular disease; chronic kidney disease; chronic liver disease; COPD; diabetes; hypertension; malignancy, including melanoma but no other skin cancers; neurologic disease; obesity; rheumatic disease; or smoking history. All diagnosis and procedure codes used to identify underlying medical conditions and COVID-19 outcomes were drawn from the FDA Sentinel Initiative COVID-19 Natural History Code List [13] . A c c e p t e d M a n u s c r i p t 7 Categorical variables were reported as counts and percentages. The prevalence of underlying medical conditions on January 20, 2020 was calculated within age-, sex-and race/ethnicity-specific subgroups. Rate ratios for a specific underlying medical condition were obtained from Poisson regression models for a COVID-19 outcome (ie, diagnosis, hospitalization, or COVID-19 disease progression outcomes while hospitalized [ICU admission, intubation and mechanical ventilation/ECMO], and death (either in or out of hospital)) within each specific subgroup, adjusting for each of the other medical conditions listed above as well as for emphysema, HIV and alcohol history. PAFs for a specific underlying medical condition were calculated as follows: where p c is the prevalence of the underlying medical condition among cases during the baseline period and RR is the adjusted rate ratio for the condition, in the specific age, sex, and race/ethnicity strata [11, 12] . PAFs, incidence rates, and adjusted rate ratios with associated 95% confidence intervals (CI) were calculated. All analyses were conducted with SAS 9.4 (SAS institute, Cary, N.C.). Demographic characteristics and prevalence of underlying medical conditions stratified by age group among persons in the cohort are shown in Across all age groups, several underlying medical conditions were associated with an increased risk for COVID-19 diagnosis and clinical outcomes of increasing severity (Figure 1, and Supplementary Table 3 ). In all but the oldest age group, cardiovascular disease and diabetes had the highest PAFs for COVID-19 diagnosis and all COVID-19 disease progression outcomes (Figure 2 In the largest study to date of the population level contribution of underlying medical conditions to COVID-19 diagnosis and COVID-19 disease progression outcomes in the United States, a large proportion of the COVID-19 disease burden was attributed to several underlying medical conditions. To our knowledge this study is the first to evaluate the population burden of COVID-19 attributable to these underlying conditions. Overall, cardiovascular disease and diabetes were associated with the highest PAFs in the study, [2, 5, 6, 9] . Neurological disease also accounted for a large proportion of the COVID-19 disease burden in persons ≥75 years, ranging from 0.05 to 0.27 depending on the outcome, consistent with previous reports that neurologic comorbidity may be an independent predictor of COVID-19 disease severity and death [15, 16] . Older age was associated with higher PAFs in this study, which is consistent with data showing poor outcomes in older patients [5, 6, 17] . PAFs for cardiovascular disease and diabetes were generally higher for female sex compared with male sex, especially among 50-64-year olds; the reasons for this are unclear. Smoking history, adjusting for medical conditions, appears to be protective for COVID-19 outcomes for <65-year-olds in this study, in contrast with results of other studies reporting worse COVID-19 outcomes for smokers [18, 19] , although it should be noted that differences in analysis (i.e., adjustment for comorbidities, age stratification) may explain the mixed findings. It has not been established whether smokers are more susceptible to SARS-CoV-2 infection; further research is needed on the role of smoking in SARS-CoV-2 infection and COVID-19 outcomes [19] and whether the relationship varies by age. Similarly, low PAFs observed for other conditions, such as hypertension, may be due to adjusting for other conditions. In the USA, cardiovascular disease is more prevalent in Black individuals, while diabetes is more common among both Black and Hispanic compared with White individuals [20] . In patients infected with COVID-19, the risk of hospitalization and mortality is higher in Black and Hispanic patients versus White and non-Hispanic patients, respectively [21] [22] [23] . could mean that these populations are routinely underdiagnosed with underlying medical conditions associated with more severe COVID-19, which may result in biased and/or underestimated estimates of the prevalence and population attributable fraction of certain medical conditions for these groups. More than 10 billion COVID-19 vaccine doses have been administered to date, and real-world evidence indicates that immunization is effective against COVID-19, including asymptomatic infection and severe disease [1, [25] [26] [27] . Studies have shown that Black, Hispanic, and Asian individuals are more likely to be vaccine hesitant, and Black individuals are less likely to be vaccinated against COVID-19 compared with White individuals [28, 29] . It is essential, therefore, that racial and ethnic disparities are addressed to minimize unequal vaccine uptake among vulnerable groups [30] . The limitations of retrospective administrative data must be considered when interpreting these results. Administrative databases rely on the accuracy and coding of diagnoses, and procedures, which can underestimate disease prevalence for conditions known to be underdiagnosed, such as obesity [31] . In this cohort, 10% of individuals were diagnosed with obesity, which is much lower than the estimated percentage of obesity in US adults of >40% [32] . A limitation of this study therefore is that it is not possible to quantify the proportion of individuals who were obese but not diagnosed with obesity, and how much this affected comorbidities such as diabetes and heart disease. Though the models were adjusted for several underlying medical conditions and stratified by age, sex and race, confounding such as from other medical conditions and socioeconomic status is possible. Although low socioeconomic status is generally associated with poor COVID-19 outcomes, study results vary [33, 34] . It is not possible to predict how analysis of A c c e p t e d M a n u s c r i p t The first COVID-19 vaccine was granted emergency use authorization on December 11, 2020 [39] . We do not have data on how many individuals, if any, in the cohort received a COVID-19 vaccine before the data cutoff of December 31, 2020. In this large retrospective cohort study, a substantial portion of the population burden of COVID-19 in the United States was attributable to two high-prevalence medical conditions: cardiovascular disease and diabetes, both of which may result from obesity [40] . Overall, 20-40% of COVID-19 M a n u s c r i p t A c c e p t e d M a n u s c r i p t 22 A c c e p t e d M a n u s c r i p t 27 Figure 2 World Health Organization. Coronavirus Disease (COVID-19) Dashboard Clinical characteristics and outcomes of hospitalised patients with COVID-19 treated in Hubei (epicentre) and outside Hubei (non-epicentre): a nationwide analysis of China Covid-19: risk factors for severe disease and death Prevalence and predictors of death and severe disease in patients hospitalized due to COVID-19: A comprehensive systematic review and meta-analysis of 77 studies and 38,000 patients Risk factors of critical & mortal COVID-19 cases: a systematic literature review and meta-analysis Epidemiological, comorbidity factors with severity and prognosis of COVID-19: a systematic review and meta-analysis Association of cardiovascular disease and 10 other pre-existing comorbidities with COVID-19 mortality: a systematic review and meta-analysis Does comorbidity increase the risk of patients with COVID-19: evidence from meta-analysis Comorbidities and multi-organ injuries in the treatment of COVID-19 World Health Organization. Frequently Asked Questions on Estimations of Attributable Burden of Disease Due to a Risk Factor Use and misuse of population attributable fractions Population attributable fraction COVID-19 Natural History Master Protocol. US Food and Drug Administration Neurological comorbidity is a predictor of death in covid-19 disease: a cohort study on 576 patients Neurological comorbidity and severity of COVID-19 The effect of smoking on COVID-19 severity: A systematic review and meta-analysis Smoking and COVID-19: What we know so far Disparities in cardiovascular disease risk in the United States Racial disparities in COVID-19 mortality across Michigan, United States. EClinicalMedicine COVID-19 Mortality in California Based on Death Certificates: Disproportionate Impacts Across Racial/Ethnic Groups and Nativity. medRxiv Disparities in COVID-19 hospitalizations and mortality among black and Hispanic patients: cross-sectional analysis from the greater Houston metropolitan area Impact and effectiveness of mRNA BNT162b2 vaccine against SARS-CoV-2 infections and COVID-19 cases, hospitalisations, and deaths following a nationwide vaccination campaign in Israel: an observational study using national surveillance data Effectiveness of COVID-19 vaccines against the B.1.617.2 variant. medRxiv Effectiveness of First Dose of COVID-19 Vaccines Against Hospital Admissions in Scotland: National Prospective Cohort Study of 5.4 Million People COVID-19 vaccine hesitancy Racial and ethnic differences in COVID-19 vaccine hesitancy and uptake. medRxiv Black-white disparities in 2009 H1N1 vaccination among adults in the United States: A cautionary tale for the COVID-19 pandemic Weighing the odds: Assessing underdiagnosis of adult obesity via electronic medical record problem list omissions. Digit Health, 2020; 6: 2055207620918715. 32. National Center for Health Statistics. Prevalence of Obesity and Severe Obesity Among Adults: United States A systematic review of racial/ethnic and socioeconomic disparities in COVID-19 Disparities in COVID-19 Outcomes by Race, Ethnicity, and Socioeconomic Status: A Systematic-Review and Meta-analysis Institute of Medicine (US) Committee on the Consequences of Uninsurance. Care Without Coverage: Too Little, Too Late. Ch 3. Effects of Health Insurance on Health National Diabetes Statistics Report. Estimates of Diabetes and Its Burden in the United States FDA Takes Key Action in Fight Against COVID-19 By Issuing Emergency Use Authorization for First COVID-19 Vaccine Obesity, Diabetes, and Cardiovascular Diseases: A Compendium chronic obstructive pulmonary disease; ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; Q, quartile, vent/ECMO, intubation and mechanical ventilation or ECMO. Only the month of death was available A c c e p t e d M a n u s c r i p t